Pesquisa Experimental em Marketing e Comportamento do Consumidor: Estado da Arte e Desafios na Era da Inteligência Artificial

Abstract Experimental research in marketing and consumer behavior is essential for understanding causal relationships and for developing strategies based on robust evidence. However, recent literature on Artificial Intelligence (AI) and Machine Learning (ML) applied to marketing reveals a predominance of conceptual approaches, systematic reviews, and non-experimental empirical studies, with scarce use of rigorous experimental designs. This paper provides a critical review of the contemporary academic literature, highlighting the lack of controlled experiments and the dominance of bibliometric methods, statistical modeling, and theoretical frameworks. It discusses the contributions of multivariate statistical methods, their integration with AI and ML, and practical examples of their application in marketing environments. The article also delves into ethical and methodological challenges such as algorithmic bias, privacy, and transparency, proposing that the advancement of the field depends on the adoption of controlled experiments combined with multivariate analyses and AI algorithms, with particular attention to social responsibility and fairness. This paper contributes by proposing an experimental research agenda that combines methodological rigor, technological innovation, and ethical reflection.